Machine learning in trading: theory, models, practice and algo-trading - page 2240

New comment
 
Vladimir Perervenko:

And how do you convert a vector to a matrix?

As my heart desires, and as I thought to implement it. I assembled from rows with a lag in the past. I reversed directions for even, odd, random lines. I tried to transform a matrix, transpose segments of a matrix. I wasn't especially inspired by the result.

 
Oleg avtomat:

That's for sure. Most people just don't know what you're talking about. Well, that's how it should be.

You have come by roundabout ways to the construction, which has been known for a long time, widely used in practice, well-proven, called "nonius tracking system".

Is that exactly what it is called?

 
elibrarius:

By the way. Is it right for us to look for patterns in different places on the chart?
I think not.
For example we found some pattern at 20 points, after which we should buy. And if this pattern was not on the 0 bar but 20-50-200 bars ago and it is too late to buy, we should sell. The flip network will find it and buy it. It will answer the question if the pattern was at the section of the chart shown to it. And we need to look for the pattern only in the right part of the chart, i.e. at the 0th bar.

So it turns out that convolutional nets are not suitable for working with the quotes. The appearance of the pattern in any place other than the 0-th bar will only impede the profitable work.
If the chart has 100 points, and the pattern has 20. Then the convolution network will give a signal that there is a pattern 80 times!

Was going to do them, but just changed my mind.

Yes, because of this I now have a window size of no more than 50 on the hour hand. Right now I'm trying to bolt on a heatmap for the activated class to see where the network found the pattern.

 
Maxim Dmitrievsky:

you can try recurrence plots. I did, but it didn't work, and it's slow, again.

Very slow))... but so far this is the only thing I've created that somehow works)))

 
Oleg avtomat:

This is one of the names. It may be called differently in different applications. But the main thing is the essence of its construction.

Look here :

https://www.google.com/search?client=opera&q=нониусная+tracking+system&sourceid=opera&ie=UTF-8&oe=UTF-8

OK, I'll take a look...

Do you think it makes sense to look for patterns in individual components ? twos, threes... ?

How would you use this tool ?

 
mytarmailS:

or series decomposition, PCA for example with inverse transformation....

you can decompose the series into atoms and put them back together


here are the first two components in the window 100

here are the second and third components

here are 3 and 4 components

here are 30 and 31 components

that way you can decompose up to 100, it's a cool thing...

all this on new data, no lag etc...


......

......

.....

Huh... Most people don't even know what I'm talking about at all )))) ))

About the usual decomposition, of course. Let the automaton tell you, he ate his passport on it and leaked 10,000 contests.
 
Maxim Dmitrievsky:
About the usual decomposition, of course. Let Von Automat tell you, he ate his passport and lost 10 000 contests on it.

You're the kind of guy who looks in a book and sees nothing.

You need more knowledge, not mere vertexes and empty certificates.

You've never understood my problem, you don't understand it, and it's unlikely you'll ever understand it.
 
Oleg avtomat:

You're the kind of guy who looks in a book and sees a fig.

you need more knowledge, not just verbs and empty certificates

my problem you have never understood, and do not understand, and it is unlikely you will ever understand

i tried to predict the sb on your tip, but it failed))

try to repeat at least one of my research (for example, in the last article) and compare it with your rattle on 3 MAs

feel, as they say, the level of empty certificates

 
Maxim Dmitrievsky:

I tried to predict SB on your tip, no luck ))

try to replicate at least one of my studies (e.g. in the last article) and compare it to your rattle on 3 MAs

feel, as they say, the level of empty certificates

You failed with SB and it turns out to be my fault !!! ;)))))))))))

I don't know what you researched, and I haven't read your articles.

All you could see was 3 MAs. Note in parenthesis, it is not MAs, but you see everything as MAs. Although what do you think are bad MAKs? Do you know what is Mashki?

In the meantime I've practically solved the problem.

 
Oleg avtomat:

You failed with SB, and it turned out to be my fault !!! ;)))))))))))

I don't know what you researched, and I haven't read your articles.

All you could see was 3 MAs. You are not able to see further.

Meanwhile, I have almost solved the problem.

I put the articles in my resume for the position of data analyst and I've received many offers from large companies

but I have another job, it's just a hobby. But if I want it, the road is open.

1...223322342235223622372238223922402241224222432244224522462247...3516
New comment
Reason: